Enterolignan-Producing Phenotypes Are Associated with Increased Gut Microbial Diversity and Altered Composition in Premenopausal Women in the United States

Published OnlineFirst December 26, 2014; DOI: 10.1158/1055-9965.EPI-14-0262

Research Article Cancer Epidemiology, Biomarkers Enterolignan-Producing Phenotypes Are & Prevention Associated with Increased Gut Microbial Diversity and Altered Composition in Premenopausal Women in the United States Meredith A.J. Hullar1, Samuel M. Lancaster1,2, Fei Li2, Elizabeth Tseng2, Karlyn Beer2, Charlotte Atkinson3, Kristiina Wah€ al€ a€4, Wade K. Copeland1, Timothy W. Randolph1, Katherine M. Newton5, and Johanna W. Lampe1,2

Abstract

Background: Lignans in plant foods are metabolized by gut ysis and regression were used to model the association between bacteria to the enterolignans, enterodiol (END) and enterolac- enterolignan excretion and the GMC. Bacteria associated with tone (ENL). Enterolignans have biologic activities important to ENL production were identified using univariate analysis and the prevention of cancer and chronic diseases. We examined the ridge regression. composition of the gut microbial community (GMC) as a con- Results: After adjusting for dietary fiber intake and adiposity, tributor to human enterolignan exposure. we found a significant positive association between ENL excretion Methods: We evaluated the association between the GMC in and either the GMC (P ¼ 0.0007), or the diversity of the GMC (P ¼ stool, urinary enterolignan excretion, and diet from a 3-day food 0.01). The GMC associated with high ENL production was distinct record in 115 premenopausal (ages 40–45 years) women in the (UNIFRAC, P < 0.003, MRPP) and enriched in Moryella spp., United States. Urinary enterolignans were measured using gas Acetanaerobacterium spp., Fastidiosipila spp., and Streptobacillus spp. chromatography–mass spectroscopy. The GMC was evaluated Conclusion: Diversity and composition of the GMC are asso- using 454 pyrosequencing of the 16S rRNA gene. Sequences were ciated with increased human exposure to enterolignans. aligned in SILVA (www.arb-silva.de). Operational taxonomic Impact: Differences in gut microbial diversity and composition units were identified at 97% sequence similarity. Taxonomic explain variation in gut metabolic processes that affect environ- classification was performed and alpha and beta diversity in mental exposures and influence human health. Cancer Epidemiol relationship to ENL production were assessed. Multivariate anal- Biomarkers Prev; 24(3); 546–54. 2014 AACR.

Introduction enzymes involved in hormone metabolism, and antitumor activities (10). High interindividual variation in excretion, circulating Epidemiologic studies have shown that the consumption of concentrations, and extent of metabolism of enterolignans exists foods of plant origin is associated with lower risk of several cancers (11). Dietary factors account for a modest amount of the variation (1). In particular, the intake of lignans, which are polyphenolic in enterolignan excretion; and often unaccounted for sources of compounds concentrated in woody portions of plants, seed coats, variation include gastrointestinal transit time, sex, and the com- and the bran layer of grains, has been inversely associated with risk position of the gut microbiome (12, 13). We hypothesize that of breast (2–7) and colon cancer (8, 9). Lignans are converted by variation in the composition of the microbiome influences the the gut microbiota to enterolignans, which are bioactive chemi- exposure of the host to lignan metabolites and that this may cals found in measurable quantities in plasma and urine. Evidence ultimately influence health outcomes. from in vitro and in vivo studies suggests that enterolignans possess Several biochemical steps are required to transform plant a variety of biologic activities relevant to human health, including lignans into enterolignans and each step is likely catalyzed by weak estrogenic and antiestrogenic properties, inhibition of consortia of bacteria that share metabolic intermediates (14). To date, no one bacteria has been identified that can completely 1Fred Hutchinson Cancer Research Center, Seattle, Washington. 2Uni- metabolize the plant lignan, secoisolarisiresinol diglucoside versity of Washington, Seattle, Washington. 3University of Bristol, (SDG) to enterolactone (ENL). For example, isolated Eggerthella Bristol, United Kingdom. 4University of Helsinki, Helsinki, Finland. lenta cannot reduce SECO; however, it can dehydroxylate 2, 3-bis- 5 Group Health Research Institute, Seattle, Washington. (3, 4-dihydroxy-benzyl) butane-1, 4-diol to enterodiol (END), Note: Supplementary data for this article are available at Cancer Epidemiology, one of the intermediary steps in ENL production (12, 15). END Biomarkers & Prevention Online (http://cebp.aacrjournals.org/). can then be converted to ENL by different bacteria (16, 17). Corresponding Author: Meredith A.J. Hullar, Fred Hutchinson Cancer Research Several more bacterial groups likely play similarly unique and Center, Division of Public Health Sciences, Cancer Prevention Program, 1100 complex biochemical roles in the transformation of plant lignans Fairview Avenue North, M4-B402, PO Box 19024, Seattle, WA 98109-1024. to enterolignans (16). Hence, the complexity and diversity Phone: 206-667-1967; Fax: 206-667-7850; E-mail: [email protected] of the gut microbial community (GMC) are essential for maxi- doi: 10.1158/1055-9965.EPI-14-0262 mizing conversion of plant lignans into enterolignans and likely 2014 American Association for Cancer Research. influences human exposure to these bacterial compounds. The

546 Cancer Epidemiol Biomarkers Prev; 24(3) March 2015

Enterolignans Associated with Increased Bacterial Diversity

objective of this study was to evaluate the association between Gut microbial community analysis GMC and urinary enterolignan excretion in a well-characterized DNA extraction. DNA was extracted from stool that had been group of premenopausal women. stored in RNAlater at 80C (20). The 16S rRNA gene was amplified and sequenced using 454 pyrosequencing primers Materials and Methods 27f and 519r (V1-V3; ref. 22) for amplicon pyrosequencing (bTE- – Research design and study participants FAP; refs. 23 27) at Research and Testing using Roche 454 FLX This observational study was conducted in premenopausal titanium instruments and reagents and following the manufac- women who were part of a larger study designed to evaluate the turer's guidelines. Sequences have been deposited in the Sequence relationship between bacterial metabolic phenotypes, diet, and Read Archive of NCBI under accession number SRP028900. biomarkers of sex steroid hormone status (18). Of the 203 women in the parent study, 120 collected a fecal sample. Of the 120 16S rRNA gene sequencing and curation. Sequences were compiled women that donated fecal samples, 116 filled out a 3-day food and processed using MOTHUR (v.1.28.0; ref. 28). Sequences were fi record (3DFR) and 101 of the samples were taken within 1 month converted to standard FASTA format from .sff les. Sequences < > of the 3DFR. One woman who provided a stool sample and 3DFR were removed if they were 300 bp, had homopolymers 8 bp, did not have an ENL measurement. Although all participants were more than one mismatch to the forward primer, more than one premenopausal, with normal menstrual cycles, we did not collect mismatch to the barcode, or ambiguous bases. Sequences were urines in conjunction with time in menstrual cycle. Previous work denoised (29), and aligned to the Silva 16S rRNA gene reference by Lampe and colleagues (19) showed no difference in END or alignment (www.arb-silva.de) using the NAST algorithm ENL by phase of cycle in a carefully controlled study of flaxseed (28, 30, 31). Sequences that did not align to the appropriate supplementation. The aims of this study are addressed in this 16S rRNA gene region were removed. Low abundance sequences subset of women. The women were recruited from Group Health, were merged to the high abundant sequences using the pre.cluster a large integrated health plan in Western Washington, and were option in MOTHUR to minimize the effect of pyrosequencing eligible to participate if they were 40 to 45 years and had under- errors in overestimating microbial diversity (32). Potentially gone a screening mammogram in the last 10 months (18). chimeric sequences were removed using ChimeraSlayer (33, 34). Women were excluded if they had more than one prescription for hormone therapy (i.e., oral contraceptives) within 18 months Analysis of the microbiome. Sequences were clustered into oper- of the sampling date; had any history of breast cancer; had breast ational taxonomic units (OTU) at 97% similarity based on the implants; had a hysterectomy or oophorectomy; used tamoxifen average neighbor-joining algorithm. The sequences were classi- fi fi or raloxifene; had any diagnosis of gastrointestinal disorders or ed using the naive Bayesian Classi er trained against an RDP fi gastrointestinal surgeries 10 years before their mammogram; or if training set as implemented in MOTHUR (27). Classi ed they had prescriptions for antibiotics, bisphosphonates, or corti- sequences were assigned to phylum and genus-level phylotypes costeroids within 3 months of their sampling date. All study (35) to characterize the community structure. To characterize the fi parameters were approved by the FHCRC and Group Health IRB alpha diversity, we used OTUs rare ed to 1,265 sequences per and all participants provided written informed consent. sample because uneven sampling depth biases diversity estimates. Diversity of the microbial community within an individual (alpha diversity), was calculated from OTUs (at 3% divergence) using the Specimen and data collection inverse Simpson index (36). Similarity in the GMC between Participants completed a health and demographics question- individuals (beta diversity) was calculated using the Theta YC naire and recorded all food and drink consumed for 3 consecutive (Q ) distance metric that accounts for shared and unshared days (18). Completed food record booklets were submitted and YC OTUs between individuals (37), and weighted and unweighted dietary intake was analyzed for nutrient content. Body composi- UniFrac (38, 39). The UniFrac approach creates a distance metric tion (% adiposity) was measured using dual energy X-ray absorp- based upon a combined phylogenetic tree and compares which tiometry (DXA; Hologic Delphi, Hologic Inc.). The fecal sample branches the two individuals have in common. We used a relaxed was collected in RNAlater (Ambion) using a method described Neighbor-Joining algorithm implemented in Clearcut (40) to previously (20). A protocol was developed to allow participants to generate the phylogenetic trees. collect samples in the privacy of their own home. The stool sample was collected in a plastic tub and a portion scooped directly into a collection tube containing approximately 5 mL of RNAlater and Statistical analysis glass beads. The sample was shaken vigorously to enhance dis- Anthropometrics, demographics, dietary, and lifestyle factors. The persal of the stool in the preservative. Specimens were immediately association of ENL and END excretion with anthropometric brought to FHCRC where they were frozen at 80 C until further measurements, demographics, and dietary and lifestyle factors fi analysis. Morning rst void urines were collected from the parti- was calculated using linear regression. cipants and frozen at 80C upon receipt until further analysis. Microbiome data cleaning. Bacterial taxa were removed if they Urinary lignan analysis represented less than 0.08% of the total sequences in a sample The first-void urine samples were analyzed for the entero- based on empirical data (background was four sequences/10,000 lignans, END and ENL, by gas chromatography–mass spectros- and 2 background) and appeared in 20% or more of the copy, with deuterated internal standards (21). All enterolignan subjects as established in the literature (41–43). The number measurements were adjusted for creatinine concentration to of sequences in each genera was converted to the relative account for urine dilution. The lowest level of quantification percentage of the total sequence abundance per individual for (LOQ) for END and ENL in 2 mL urine was 70 mg/L. multivariate analysis.

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 547

Hullar et al.

Figure 1. The characterization of the GMC in a cross-sectional analysis of premenopausal women. The ﬁrst three axes of the NMS analysis of the GMC account for up to 85% of the variation in the data. The vectors radiating from the centroid and overlain on the NMS plot represent the relative association of the genera of bacteria and the axes (r2 > 0. 49) and the magnitude of the association. Cluster 1 (Bacteroides); Cluster 2 (Prevotella); Cluster 3 (Dethiosulﬁtobacter, Pyramidobacter, Oscillibacter).

Multivariate analysis. Distance estimates between each pair of ized by cluster analysis from a Qyc based distance matrix and samples were calculated using the Jensen–Shannon Divergence clustered using the unweighted pair group method with arithme- (JSD; refs. 44, 45). Data dimension reduction was performed tic mean (Fig. 2). using nonmetric multidimensional scaling (NMS) on the matrix Statistical analyses were implemented in PC-ORD version of JSD distances generated between each pair of samples (44, 45). 6.14, vegan (49), and R version 3.1 and visualized using iTOL A joint plot was used to visualize the relative strength and 2.1 (50). magnitude of the association between each genera and the NMS axes (r2 > 0.49; Fig. 1; ref. 46). To model the variation in ENL Results excretion and the microbiome, the NMS axes used to describe the Anthropometrics, demographics, diet, and phenotype microbial community were included in separate models of the The mean age of participants (n ¼ 115) was 42 years, and GMC (axis 1, axis 2, or axis 3) and anthropometric, dietary, and the majority were white, with at least some college education lifestyle factors. Three models incorporating each NMS axes (Table 1). All women had urinary ENL concentrations above the separately and using urinary ENL as the response variable were LOQ (>70 mg/L); mean (SD) urinary ENL was 3.08 4.44 mg/ investigated. mg creatinine. Fifty-eight participants had urinary END concentrations above the detection limit (>70 mg/L); mean END in these women was 0.23 0.53 mg/mg creatinine (Table 1). Between the Identification of bacterial enterotypes. To investigate whether the highest and lowest tertiles of ENL excretion, there were significant gut microbiome of our study participants clustered by enterotype, decrease in body mass index (BMI; P < 0.003) and adiposity (P < the JSD (44) was computed on pairs of samples followed by 0.004), and a significant increase in dietary fiber intake (P ¼ clustering using partitioning around medoids (PAM) clustering. 0.0008), education (P ¼ 0.03), and self-reported frequency of Optimal numbers of clusters were determined using the Calinski– diarrhea (P ¼ 0.03; Table 1). Significant increases in dietary fiber Harabasz index (47). intake (P < 0.002), ENL excretion (P ¼ 0.003), and education (P ¼ 0.0004) were observed between participants above and below the Identification of bacterial genera associated with ENL production. LOQ for urinary END (data not shown). To reveal which, if any, taxa were associated with ENL production, we used two regression approaches. The first approach was a univariate regression model in which each taxon was considered Evaluation of gut microbial community individually. The second approach was a penalized regression Using 454 pyrosequencing of the V1-V3 region, a total of 1.4 approach in which all of the taxa were considered simultaneously million raw sequences were processed. The resulting pool of (see Supplementary Data; ref. 48). OTUs associated with ENL 644,956 sequences, which averaged 5,201 2,741 sequences per production using both regression approaches were considered the participant, was analyzed. The sequences were, on average, 367 most parsimonious. 29 bp long. The trimmed sequences represented a total of 342 bacterial genera (phylotypes), of which, 133 OTUs met the Microbial diversity and ENL production. ANOVA was used to test microbiome data cleaning criterion (see Materials and Methods). the null hypothesis that there was no statistical difference in the We used this multivariate matrix containing 133 OTU's for alpha diversity of the GMC between high or low ENL excreters statistical analysis. based on tertiles of excretion. Multiple response permutation Bacteria were distributed across phyla: Firmicutes (68%), Bac- procedure (MRPP) was used to test the null hypothesis that there teroidetes (27%), Proteobacteria (2%), Synergistes (1.0%), Acti- was no statistical difference in the composition of the GMC nobacteria (0.5%), Fusobacterium (0.4%), Verrucomicrobia between high or low ENL excreters based on tertiles of excretion. (0.02%), Ternicutes (0.1%), and Lentisphaerae (0.05%). Good's Differences in the GMC by tertiles of ENL excretion were visual- coverage was 0.996 0.003.

548 Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 Cancer Epidemiology, Biomarkers & Prevention

Enterolignans Associated with Increased Bacterial Diversity

Figure 2. The composition of the GMC is signiﬁcantly different between high and low ENL excreters (MRPP; P < 0.001). Cluster analysis of beta diversity (Qyc) of the microbiome in tertiles of ENL excretion. Bars are the amount of urinary ENL (mg/mg creatinine) color-coded for tertile; low, light gray; medium, white; and high, black. The three samples without bars are technical replicates.

NMS on the matrix of JSD distances was used to describe the energy intake (P ¼ 0.02), and significantly negatively associated microbiome associated with variation in ENL excretion with the percentage of adiposity (P ¼ 0.02). ENL production was (44, 46, 51). The final solution for NMS analysis of the GMC also significantly positively associated with microbial alpha diver- patterns had a stable stress value of 13.32, after 400 iterations sity (P ¼ 0.01), and fiber intake adjusted for energy intake (P ¼ using a random seed of 2564. The three axes cumulatively 0.01). In contrast, there was no significant association between accounted for 87% of the variation in the GMC measured using END excretion and the GMC composition or diversity (data not 16S rRNA gene data; axes 1, 2, and 3 accounted for 35, 13, and shown). 39%, respectively (Fig. 1). Bacteroides was negatively correlated To reveal individual OTU that were associated with ENL, we with axis 3 (r ¼0.79), Prevotella was negatively correlated with additionally considered two regression approaches. Using either axis 1 (r ¼0.82), and Oscillibacter (r ¼ 0.94), Dethiosulfatibacter univariate or Ridge regression, we identified four bacterial genera, (r ¼ 0.86), and Pyramidobacter (r ¼ 75) were positively correlated Moryella (52), Acetanaerobacterium (53), Fastidiosipila (54), and with axis 3 (Fig. 1). We also identified three clusters in the Streptobacillus (55, 56) that were significantly increased in the high microbial community (Fig. 1). Each cluster was subsequently ENL excreter tertile (see Supplementary Data). Genera associated observed to be associated with different dominant microbial with ENL excretion represented between 0.03% and 0.7% of the genera: (i) Bacteroides, (ii) Prevotella, or (iii) a combination of total microbiome (Table 2; see Supplementary Material for more Pyramidobacter, Dethiosulfatibacter, and Oscillibacter (Supplemen- details). tary Figs. S1, S2, and S3). The distribution of the relative percent- Bacterial diversity was positively associated with high ENL age of these groups showed discrete grouping for some of the excretion. The bacterial alpha diversity, a measure of the variation dominant genera (Supplementary Fig. S3). The NMS axes were of the composition of the GMC within a person, was significantly used in regression models to relate the GMC to ENL excretion and different between women in ENL excreter groups (Inverse Simp- dietary intake. son index, ANOVA, F ¼ 12.90, n ¼ 115, P < 0.0001). More We fit a linear regression model for the association of ENL specifically, there was a significant difference in the alpha diversity excretion with NMS axis 1 (JSD), adiposity, and fiber (tertiled, between low and high (Tukey, q ¼ 6.5, P < 0.0001) excreters. Beta using the third tertile as reference) and adjusted for calories (n ¼ diversity, a comparison of bacterial diversity between subjects, 101). ENL excretion was significantly positively associated with was significantly different in women in the low ENL group as GMC described by NMS axis 1 (P ¼ 0.0007), and fiber adjusted for compared with those the high ENL group using weighted Unifrac,

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 549

Hullar et al.

Table 1. Study participant anthropometrics, demographics, and diet and lifestyle factors by tertiles of ENL excretion First tertile Second tertile Third tertile n ¼ 38 n ¼ 38 n ¼ 39 P Continuous descriptive statistics, mean (SD) Urinary ENL (mg/mg creatinine; ENL) 0.46 (0.37) 1.86 (0.91) 5.66 (4.52) NA Urinary END (mg/mg creatinine; END) 0.06 (0.13) 0.09 (0.13) 0.56 (1.33) 0.35 Age, y 42.49 (1.41) 42.05 (1.35) 42.59 (1.32) 0.78 BMI (kg/m2) 27.10 (4.18) 25.61 (4.56) 24.82 (5.2) 0.003 Adiposity (% fat) 35.33 (5.68) 34.82 (7.19) 30.85 (7.07) 0.004 Energy (kcal/d) 1,937.22 (449.07) 1,894.27 (390.92) 1,956.14 (395.72) 0.84 Carbohydrate (g/d) 230.18 (67.32) 220.59 (53.48) 245.65 (62.93) 0.38 Protein (g/d) 78.02 (19.54) 76.57 (15.81) 79.07 (19.73) 0.81 Fat (g/d) 77.09 (25.22) 76.44 (19.85) 72.24 (21.2) 0.32 Dietary ﬁber (g/d) 18.51 (6.4) 18.12 (6.1) 24.40 (8.87) 0.0008 Bowel movements (n/week) 8.76 (4.83) 7.14 (4.11) 7.22 (2.5) 0.59 Categorical descriptive statistics, n (%) Education, n (%) 12 years 1 (2.7%) 4 (11.11%) 1 (2.7%) 13–16 years 26 (70.27%) 20 (55.56%) 15 (40.54%) 17þ years 10 (27.03%) 12 (33.33%) 21 (56.76%) 0.03 Residence at birth, n (%) Rural 9 (25.71%) 6 (18.75%) 13 (34.21%) Not rural 26 (74.29%) 26 (81.25%) 25 (65.79%) 0.34 Ethnicity, n (%) White 28 (82.35%) 29 (90.62%) 34 (94.44%) Asian 4 (11.76%) 2 (6.25%) 0 (0%) Other 2 (5.88%) 1 (3.12%) 2 (5.56%) 0.29 Breast fed as an infant, n (%) Yes 17 (58.62%) 12 (42.86%) 20 (58.82%) No 12 (41.38%) 16 (57.14%) 14 (41.18%) 0.37 Self-reported IBS, n (%) Yes 2 (5.71%) 3 (9.09%) 3 (7.89%) No 33 (94.29%) 30 (90.91%) 35 (92.11%) 0.90 Diarrhea, n (%) Yes 15 (42.86%) 12 (36.36%) 6 (15.79%) No 20 (57.14%) 21 (63.64%) 32 (84.21%) 0.03 Laxative use, n (%) Yes 3 (8.57%) 0 (0%) 1 (2.63%) No 32 (91.43%) 33 (100%) 37 (97.37%) 0.21 Constipation, n (%) Yes 7 (20%) 12 (36.36%) 10 (26.32%) No 28 (80%) 21 (63.64%) 28 (73.68%) 0.35 Abbreviation: IBS, irritable bowel syndrome.

which normalizes for the number of sequences in an OTU cluster excretion in both ENL and END (Table 1). END, an intermediate (MRPP; A ¼ 0.014, P ¼ 0.003; 999 permutations), unweighted compound in the conversion of some plant lignans to ENL, was Unifrac (MRPP; A ¼ 0.005, P ¼ 0.001; 999 permutations) or using 10-fold lower than ENL in our study participants. Enterolignan the Qyc (MRPP; A ¼ 0.02, P < 0. 001; 999 permutations) as excretion (END þ ENL) ranged from 0 to 30 mg/mg creatinine. visualized in Fig. 2. These ranges encompass or are higher than other study populations consuming a predominantly western diet (57–59). The metabolism of lignans in the gut occurs by a consortia of Discussion microorganisms through a series of reactions (60–64), and can In this cross-sectional study, we evaluated differences in result in a measurable amount of bacterial metabolites in host GMC in relation to lignin-metabolizing phenotypes. We found systemic circulation. There was a significant difference in the that GMC differed by tertile of ENL but not END excretion. composition of the microbiome between the highest and lowest GMC diversity increased with greater ENL excretion. We iden- tertiles of ENL excretion (Fig. 2). Genera representing median tified components of the microbiome associated with excretion values between 0.03% and 0.34% of the microbiome were asso- of ENL. These data suggest that the environmental exposures ciated with ENL production (Table 2). Although each bacterial from dietary intake can be altered by the metabolic capacity of population may represent a minor component of the micro- the more minor components of the gut microbiome to influ- biome, when considered across the sum of the entire metabolic ence health outcomes. transformation from plant lignans to enterolignans, they had a Pharmacokinetic studies have shown a wide variation in enter- measurable impact on ENL excretion. These findings are in olignan excretion, in both magnitude and time of excretion (13). keeping with those of Clavel and colleagues (12) who reported These excretion patterns have been shown to vary within different that in vitro lignan degradation was associated with minor com- human populations. In our study, we found a wide range of ponents (<1%) of the gut microbiome.

550 Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 Cancer Epidemiology, Biomarkers & Prevention

Enterolignans Associated with Increased Bacterial Diversity

Table 2. List of genera whose abundances exhibit an association with ENL excretion in 115 women Na OTUb Low High Univariate estimates Acetanaerobacterium 100 0.1850 0.0006 0.0063 Acetitomaculum 29 0.0190 0.0006 0.0069 Acetivibrio 99 0.3990 0.0004 0.0063 Bacteroides 115 13.7820 0.0064 0.0002 Cerasibacillus 43 0.1610 0.0006 0.0221 Dethiosulfatibacter 76 1.1360 0.0039 0.0113 Ethanoligenens 64 0.0660 0.0016 0.0101 Eubacterium 84 0.1010 0.0016 0.0103 Fastidiosipila 109 0.3420 0.0007 0.0099 Holdemania 74 0.0440 0.0200 0.0012 Moryella 68 0.0320 0.0055 0.0147 Pseudobutyrivibrio 115 9.6760 0.0071 0.0021 Pyramidobacter 81 0.7040 0.0018 0.0117 Reichenbachiella 77 0.5940 0.0002 0.0067 Ruminococcus 96 0.2900 0.0010 0.0083 Sarcina 27 0.0370 0.0011 0.0064 Sedimentibacter 56 0.0380 0.0029 0.0101 Streptobacillus 70 0.3260 0.0000 0.0137 Synergistes 26 0.1000 0.0006 0.0044 Victivallis 38 0.1010 0.0015 0.0140 Ridge regression Acetanaerobacterium 100 0.185 0.0009 0.0076 Butyricimonas 61 0.112 0.0019 0.0076 Clostridium 77 0.072 0.0006 0.0066 Coprococcus 115 10.836 0.0093 0.0518 Fastidiosipila 109 0.342 0.0001 0.0101 Moryella 68 0.032 0.0001 0.0010 Oscillibacter 115 2.934 0.0015 0.0388 Prevotella 54 0.639 0.0002 0.0421 Sharpea 29 0.161 0.0016 0.0060 Streptobacillus 70 0.326 0.0036 0.0174 NOTE: The list includes OTU that exhibit an association with ENL by way of a conﬁdence interval that does not include zero in either a univariate model (upper part of table) or a multivariate model (lower part of table); four OTUs appear in both. Underlined genera were common to both methods. aNumber of subjects for whom a non-zero OTU abundance was observed. bMedian of the relative abundance of each OTU (100).

In humans, the microbiome plays an essential role in the been associated with obesity (69), reduced cognitive function in catabolism of dietary fibers because the human genome does not the elderly (41), Clostridium difficile–associated disease (70), and encode the range of enzymes required to degrade the biochemical irritable bowel disease (26). structural diversity found in plant materials. We found that Dietary intake of lignans is a major factor that influences microbial diversity was significantly different (Simpson index, variation in urinary ENL. A high plant-food diet, rich in whole P < 0.05) between high and low ENL excreters and was positively grains, nuts, seeds and fruits, and vegetables, has been associated associated with fiber intake. The association between ENL excre- with higher production of enterolignans (71). We found that tion and microbial diversity may reflect the complexity of the dietary fiber intake was significantly associated with ENL produc- microbial metabolism involved in ENL metabolism because there tion (Table 1). This has been observed in previous studies and are several transformations involved in the production of enter- supports the findings that lignan content and dietary fiber content olignans from lignans and different bacteria species can catalyze of foods are often highly correlated (72–74). In vitro incubations each step (15–17, 65–67). In a recent study comparing human have also shown that the type of fiber may influence ENL pro- populations consuming either a western diet or a rural sub- duction (61). Insoluble fiber includes lignan-rich categories, such Saharan traditional diet, dietary fiber intake was also associated as seed coats and bran layers. Higher rates of ENL production were with gut microbial diversity (68). The association between intake associated with insoluble fiber in other cross-sectional studies of dietary fiber as plant material and gut microbiome diversity (75). may not be apparent because most studies of the human micro- Obesity in adults has been linked to the gut microbiome, and biome have been conducted in human populations consuming a the altered functional potential of the obesidogenic microbiome western diet, which is traditionally low in fiber. Furthermore, influences myriad negative health outcomes (76). Consistent with microbial functional gene diversity may be associated with long- previous studies (77, 78), we found that adiposity was inversely term dietary patterns that include a high fruit and vegetable intake, associated with ENL production even after controlling for fiber and therefore a higher fiber intake (69). In addition to being intake. Kilkkinen and colleagues (78) found an inverse associa- associated with a beneficial phenotype, high microbial diversity tion between ENL production and BMI. More specifically, they associated with dietary fiber intake may provide a key to main- found that normal weight women produced significantly higher taining resilience of the host to infection and other environmental ENL than their underweight or obese counterparts. Frankenfeld impacts. For example, dietary patterns that were associated with (11) found that overweight and obese individuals were less likely decreased microbiome structural and functional diversity have to excrete high levels of ENL. This association was potentially

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 551

Hullar et al.

stronger in women than in men. In our study, the inverse asso- this, the ranges of enterolignan excretion encompass or were ciation between adiposity and ENL production could reflect the slightly higher to those found in predominantly white popula- lower intake of high-fiber foods (Table 1), and therefore a lower tions and data associated with lignan metabolism by gender are intake of plant lignans by women with a higher percentage of equivocal (58, 59, 72, 73, 94, 95). Dietary data were not collected body fat. at the same time as the urine for enterolignan phenotyping or as Although ENL is produced by bacterial consortia, identification the stool sample. However, to optimize sample size and minimize of the bacteria involved in ENL production has previously been lag time in sampling, we excluded participants whose 3DFR was based on isolation of a pure culture of the organism associated taken greater than 1 month apart from stool samples. Our study is with a specific part of the metabolic pathway (16, 66, 79). We also limited by the fact that participants were consuming their found four bacterial genera (Moryella, Streptobacillus, Fastidiosipila, habitual diets, which contributed to variation in the amounts and and Acetanaerobacterium) associated with the high ENL producers types of plant lignans consumed, and therefore also contributed (Table 2). Although they have not been identified before in ENL to the variation in END and ENL excretion. production (80), the bacteria we identified are closely related to The gut microbiome can influence the magnitude and flux of genera that have been previously associated with the bacterial dietary metabolites to which the host is exposed. Major bacterial metabolism of lignans. For example, the initial production of parameters to consider in enterolignan bioavailability include secoisolariciresinol from secoisolariciresinol diglucoside is asso- how dietary fiber influences diversity, community composition, ciated with glucosidase activity. Bacteria related to the genera and functional activity of the microbiome that appears to be Streptobacillus produce extracellular enzymes involved in glucose altered in the metabolic phenotypes studied here. We observed cleavage from complex molecules (55, 56), as do the genera differences in GMC in relation to ENL excretion in a group of Fastidiosipilia (54). Once these sugars are made available, mem- premenopausal women. Bacterial diversity and community struc- bers of the genera Acetanaerobacterium and Moryella are able to ture were significantly associated with ENL excretion, and we ferment glucose to acetate or butyrate (52, 53). Also, Clostridia identified several bacterial groups newly associated with in vivo closely related to Oscillibacter are involved in anaerobic ring ENL production. Future studies of the microbial response to diet cleavage and ring cleavage of methoxylated compounds (81, 82). will further our understanding of how environmental exposures Diet can influence the gut microbiome in humans (69, 76, 83– may be altered by the gut microbiome and influence health 86). Clusters dominated by Bacteroides (Enterotype 1) or Prevotella outcomes. (Enterotype 2) have been associated with diets rich in fats and carbohydrates (51, 87, 88). In our study of healthy women, the Disclosure of Potential Conflicts of Interest bacterial composition of groups 1 and 2 is similar to previous No potential conflicts of interest were disclosed. findings (51), although the composition of a third group of bacteria varied from other published report (Fig. 1; Supplemen- Authors' Contributions tary Fig. S2; refs. 89–91). Enterotype groupings have been iden- Conception and design: M.A.J. Hullar, C. Atkinson, J.W. Lampe tified in several studies, but studies have found either fewer than Development of methodology: M.A.J. Hullar, F. Li, K.M. Newton three enterotypes (86) or no pattern (92) when considering Acquisition of data (provided animals, acquired and managed patients, fi provided facilities, etc.): M.A.J. Hullar, F. Li, C. Atkinson, K.M. Newton, adults consuming western diets. A classi cation system based J.W. Lampe upon functional genes instead of the dominant members of the Analysis and interpretation of data (e.g., statistical analysis, biostatistics, gut microbiome may be more appropriate because they do not computational analysis): M.A.J. Hullar, S.M. Lancaster, F. Li, E. Tseng, K. Beer, necessarily reflect the complexity of metabolism catalyzed by W.K. Copeland, T.W. Randolph microbial consortia involved in phytochemical metabolism that Writing, review, and/or revision of the manuscript: M.A.J. Hullar, S.M. Lan- € € € influences health (51, 93). caster, C. Atkinson, K. Wahala, W.K. Copeland, T.W. Randolph, K.M. Newton, J.W. Lampe This study has several strengths. This study was conducted in a Administrative, technical, or material support (i.e., reporting or organizing sample of premenopausal women selected originally for a study data, constructing databases): F. Li of isoflavone metabolism and hormonal factors (18). As a result, Other (contributed to chemical matters): K. W€ah€al€a factors that could affect GMC (e.g., antibiotic use) were considered in participant selection. Dietary intake was measured using a Grant Support 3DFR rather than relying on a food frequency questionnaire. Body This work was supported by NIH grants R01 CA97366 (J.W. Lampe, PI), U01 fat was measured using DXA, providing a more accurate measure CA63731 (E. White, PI), U54 CA116847(A. McTiernan, PI), and R03 CA115209 of adiposity. High-throughput sequencing was used to character- (J.W. Lampe, PI), and Kellogg Corporate Citizens Fund and Fred Hutchinson Cancer Research Center. ize the GMC, which, given that ENL is produced by metabolic The costs of publication of this article were defrayed in part by the payment of consortia of bacteria, was able to capture the complexity of the page charges. This article must therefore be hereby marked advertisement in GMC–ENL association. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Our study has some weaknesses. Most of the women recruited in this study were white, well educated, and were recruited to achieve a wide range of breast density. Therefore, the findings may Received March 18, 2014; revised December 5, 2014; accepted December 9, not be applicable to the general population. However, despite 2014; published OnlineFirst December 26, 2014.

References 1. World Cancer Research Fund. Food, nutrition, physical activity, and the 2. Buck K, Zaineddin AK, Vrieling A, Heinz J, Linseisen J, Flesch-Janys D, et al. prevention of cancer: a global perspective. Washington, DC: American Estimated enterolignans, lignan-rich foods, and ﬁbre in relation to survival Institute for Cancer Research; 2007. after postmenopausal breast cancer. Br J Cancer 2011;105:1151–7.

552 Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 Cancer Epidemiology, Biomarkers & Prevention

Enterolignans Associated with Increased Bacterial Diversity

3. Buck K, Zaineddin AK, Vrieling A, Linseisen J, Chang-Claude J. Meta- 25. Wickham ME, Lupp C, Vazquez A, Mascarenhas M, Coburn B, Coombes analyses of lignans and enterolignans in relation to breast cancer risk. Am BK, et al. Citrobacterium rodentium virulence in mice associates with bac- J Clin Nutr 2010;92:141–53. terial load and the type III effector NleE. Microbes Infect 2007;9:400–7. 4. Guglielmini P, Rubagotti A, Boccardo F. Serum enterolactone levels and 26. Willing BP, Dicksved J, Halfvarson J, Andersson AF, Lucio M, Zheng Z, et al. mortality outcome in women with early breast cancer: a retrospective A pyrosequencing study in twins shows that gastrointestinal microbial cohort study. Breast Cancer Res Treat 2012;132:661–8. profiles vary with inflammatory bowel disease phenotypes. Gastroenterol 5. McCann SE, Hootman KC, Weaver AM, Thompson LU, Morrison C, Hwang 2010;139:1844–54. H, et al. Dietary intakes of total and specific lignans are associated with 27. Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classifier for rapid clinical breast tumor characteristics. J Nutr 2012;142:91–8. assignment of rRNA sequences into the new bacterial taxonomy. Appl 6. Toure A, Xueming X. Flaxseed lignans: source, biosynthesis, metabolism, Environ Microbiol 2007;73:5261–7. antioxidant activity, bio-active components, and health benefits. Compr 28. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, et al. Rev Food Sci Food Saf 2010;9:261–9. Introducing MOTHUR: open-source, platform-independent, community- 7. Velentzis LS, Cantwell MM, Cardwell C, Keshtgar MR, Leathem AJ, Wood- supported software for describing and comparing microbial communities. side JV. Lignans and breast cancer risk in pre- and post-menopausal women: Appl Environ Microbiol 2009;75:7537–41. meta-analyses of observational studies. Br J Cancer 2009;100:1492–8. 29. Quince C, Lanzen A, Davenport RJ, Turnbaugh PJ. Removing noise from 8. Cotterchio M, Boucher BA, Manno M, Gallinger S, Okey A, Harper P. pyrosequenced amplicons. BMC Bioinformatics 2011;12:38–56. Dietary phytoestrogen intake is associated with reduced colorectal cancer 30. Schloss PD. Secondary structure improves OTU assignments of 16S rRNA risk. J Nutr 2006;136:3046–53. gene sequences. ISME J 2013;7:457–60. 9. Kuijsten A, Arts IC, Hollman PC, van't Veer P, Kampman E. Plasma 31. Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig WG, Peplies J, et al. SILVA: enterolignans are associated with lower colorectal adenoma risk. Cancer a comprehensive online resource for quality checked and aligned ribo- Epidemiol Biomarkers Prev 2006;15:1132–6. somal RNA sequence data compatible with ARB. Nucleic Acids Res 10. Webb AL, McCullough ML. Dietary lignans: potential role in cancer 2007;35:7188–96. prevention. Nutr Cancer 2005;51:117–31. 32. Huse SM, Welch DM, Morrison HG, Sogin ML. Ironing out the wrinkles in 11. Frankenfeld CL. Relationship of obesity and high urinary enterolignan the rare biosphere through improved OTU clustering. Environ Microbiol concentrations in 6806 children and adults: analysis of National Health 2010;12:1889–98. and Nutrition Examination Survey data. Eur J Clin Nutr 2013;67:887–9. 33. Edgar RC, Haas BJ, Clemente JC, Quince C, Knight R. UCHIME improves 12. Clavel T, Henderson G, Engst W, Dore J, Blaut M. Phylogeny of human sensitivity and speed of chimera detection. Bioinformatics 2011;27: intestinal bacteria that activate the dietary lignan secoisolariciresinol 2194–200. diglucoside. FEMS Microbiol Ecol 2006;55:471–8. 34. Haas BJ, Gevers D, Earl AM, Feldgarden M, Ward DV, Giannoukos G, et al. 13. Kuijsten A, Arts IC, Vree TB, Hollman PC. Pharmacokinetics of entero- Chimeric 16S rRNA sequence formation and detection in Sanger and 454- lignans in healthy men and women consuming a single dose of secoiso- pyrosequenced PCR amplicons. Genome Res 2011;21:494–504. lariciresinol diglucoside. J Nutr 2005;135:795–801. 35. Kim M, Morrison M, Yu Z. Evaluation of different partial 16S rRNA gene 14. Heinonen S, Nurmi T, Liukkonen K, Poutanen K, W€ah€al€aK,DeyamaT,etal. sequence regions for phylogenetic analysis of microbiomes. J Microbiol In vitro metabolism of plant lignans: new precursors of mammalian lignans Methods 2011;84:81–7. enterolactone and enterodiol. J Agric Food Chem 2001;49:3178–86. 36. Magurran A. Measuring biological diversity. Oxford, UK: Blackwell; 2004. 15. Jin JS, Zhao YF, Nakamura N, Akao T, Kakiuchi N, Min BS, et al. Enantio- 37. Yue JC, Clayton MK. A similarity measure based on species proportions. selective dehydroxylation of enterodiol and enterolactone precursors by Commun Stat-Theor M 2005;34:2123–31. human intestinal bacteria. Biol Pharm Bull 2007;30:2113–9. 38. Lozupone C, Knight R. UniFrac: a new phylogenetic method for comparing 16. Clavel T, Lippman R, Gavini F, Dore J, Blaut M. Clostridium sacchar- microbial communities. Appl Environ Microbiol 2005;71:8228–35. ogumia sp. nov.andLactonifactor longoviformis gen. nov., sp. nov. two 39. Lozupone CA, Hamady M, Kelley ST, Knight R. Quantitative and qualita- novel human faecal bacteria involved in the conversion of the dietary tive beta diversity measures lead to different insights into factors phytoestrogen secoisolariciresinol diglucoside. Syst Appl Microbiol that structure microbial communities. Appl Environ Microbiol 2007; 2007;30:16–26. 73:1576–85. 17. Jin JS, Hattori M. Human intestinal bacterium, Strain END-2, is responsible 40. Sheneman L, Evans J, Foster JA. Clearcut: a fast implementation of relaxed for demethylation as well as lactonization during plant lignan metabolism. neighbor joining. Bioinformatics 2006;22:2823–4. Biol Pharm Bull 2010;33:1443–7. 41. Claesson MJ, Cusack S, O'Sullivan O, Greene-Diniz R, de Weerd H, 18. Atkinson C, Newton KM, Bowles EJ, Yong M, Lampe JW. Demographic, Flannery E, et al. Composition, variability, and temporal stability of the anthropometric, and lifestyle factors and dietary intakes in relation to intestinal microbiota of the elderly. Proc Natl Acad Sci U S A 2011;108 daidzein-metabolizing phenotypes among premenopausal women in the Suppl 1:4586–91. United States. Am J Clin Nutr 2008;87:679–87. 42. Kuczynski J, Lauber CL, Walters WA, Parfrey LW, Clemente JC, Gevers D, 19. Lampe JW, Martini MC, Kurzer MS, Adlercreutz H, Slavin JL. Urinary lignan et al. Experimental and analytical tools for studying the human micro- and isoflavonoid excretion in premenopausal women consuming flaxseed biome. Nat Rev Genet 2012;13:47–58. powder. Am J Clin Nutr 1994;60:122–8. 43. Navas-Molina JA, Peralta-Sanchez JM, Gonzalez A, McMurdie PJ, Vazquez- 20. Li F, Hullar MA, Schwarz Y, Lampe JW. Human gut bacterial communities Baeza Y, Xu ZJ, et al. Advancing our understanding of the human micro- are altered by addition of cruciferous vegetables to a controlled fruit- and biome using QIIME. Methods Enzymol 2013;531:371–444. vegetable-free diet. J Nutr 2009;139:1685–91. 44. Endres DM, Schindelin JE. A new metric for probability distributions. IEEE 21. Atkinson C, Lampe JW, Scholes D, Chen C, W€ah€al€a K, Schwartz SM. Lignan Trans Info Theory 2003;49:1858–60. and isoflavone excretion in relation to uterine fibroids: a case–control study 45. Kruskal JB. Nonmetric multidimensional scaling:a numerical method. of young to middle-aged women in the United States. Am J Clin Nutr Psycometrika 1964;29:115–29. 2006;84:587–93. 46. McCune B, Grace JB. Analysis of ecological communities. First edition. 22. Li F, Hullar MA, Lampe JW. Optimization of terminal restriction fragment Glendenen Beach, OR: MjM Software; 2002. polymorphism (TRFLP) analysis of human gut microbiota. J Microbiol 47. Calinski T, Harabasz J. A dendrite method for cluster analysis. Comm Stat Methods 2007;68:303–11. 1974;3:1–27. 23. Dowd SE, Sun Y, Wolcott RD, Domingo A, Carroll JA. Bacterial tag-encoded 48. Brumback BA, Ruppert D, Wand MP. Variable selection and function FLX amplicon pyrosequencing (bTEFAP) for microbiome studies: bacterial estimation in additive nonparametric regression using a data-based pri- diversity in the ileum of newly weaned Salmonella-infected pigs. Food- or—comment. J Am Stat Assoc 1999;94:794–7. borne Pathog Dis 2008;5:459–72. 49. Dixon P. VEGAN, a package of R functions for community ecology. 24. Callaway TR, Dowd SE, Wolcott RD, Sun Y, McReynolds JL, Edrington TS, J Vegetation Sci 2003;14:927–30. et al. Evaluation of the bacterial diversity in cecal contents of laying hens fed 50. Letunic I, Bork P. Interactive Tree Of Life v2: online annotation and various molting diets by using bacterial tag-encoded FLX amplicon pyr- display of phylogenetic trees made easy. Nucleic Acids Res 2011;39: osequencing. Poult Sci 2009;88:298–302. W475–8.

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 553

Hullar et al.

51. Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende DR, 72. Horner NK, Kristal AR, Prunty J, Skor HE, Potter JD, Lampe JW. Dietary et al. Enterotypes of the human gut microbiome. Nature 2011;473: determinants of plasma enterolactone. Cancer Epidemiol Biomarkers Prev 174–80. 2002;11:121–6. 52. Carlier JP, K'Ouas G, Han XY. Moryella indoligenes gen. nov., sp. nov., an 73. Lampe JW, Gustafson DR, Hutchins AM, Martini MC, Li S, W€ah€al€aK,etal. anaerobic bacterium isolated from clinical specimens. Int J Syst Evol Urinary isoflavonoid and lignan excretion on a Western diet: relation to Microbiol 2007;57:725–9. soy, vegetable, and fruit intake. Cancer Epidemiol Biomarkers Prev 53. Chen S, Dong X. Acetanaerobacterium elongatum gen. nov., sp. nov., 1999;8:699–707. from paper mill waste water. Int J Syst Evol Microbiol 2004;54: 74. Lampe JW. Isoflavonoid and lignan phytoestrogens as dietary biomarkers. 2257–62. J Nutr 2003;133:956S-64S. 54. Falsen E, Collins MD, Welinder-Olsson C, Song Y, Finegold SM, Lawson 75. Aura A-M, Myllym€aki O, Bailey M, Penalvo JL, Adlercreutz H, Poutanen K. PA. Fastidiosipila sanguinis gen. nov., sp. nov., a new Gram-positive, Interrelationships between carbohydrate type, phenolic acids and initial coccus-shaped organism from human blood. Int J Syst Evol Microbiol pH on in vitro conversion of enterolactone from rye lignans. In:Salovaara H, 2005;55:853–8. Gates F, Tenkanen M, editors. Dietary fibre: components and functions. 55. Larcia LH, Estacio RC, Dalmacio LM. Bacterial diversity in Philippine Wageningen, The Netherlands: Wageningen Academic Publishers; 2007. fermented mustard (Burong mustasa) as revealed by 16S rRNA gene analysis. p. 235–45. Beneficial Microbes 2011;2:263–71. 76. Hullar MA, Lampe JW. The gut microbiome and obesity. Nestle Nutr Inst 56. Elliott SP. Rat bite fever and Streptobacillus moniliformis. Clin Microbiol Workshop Ser 2012;73:67–79. Rev 2007;20:13–22. 77. Frankenfeld CL, Lampe JW, Shannon J, Gao DL, Li W, Ray RM, et al. Fruit 57. Eichholzer M, Richard A, Nicastro HL, Platz EA, Linseisen J, Rohrmann S. and vegetable intakes in relation to plasma nutrient concentrations in Urinary lignans and inflammatory markers in the US National Health and women in Shanghai, China. Public Health Nutr 2012;15:167–75. Nutrition Examination Survey (NHANES) 1999–2004 and 2005–2008. 78. Kilkkinen A, Valsta LM, Virtamo J, Stumpf K, Adlercreutz H, Pietinen P. Cancer Causes Control 2014;25:395–403. Intake of lignans is associated with serum enterolactone concentration in 58. Sun Q, Wedick NM, Pan A, Townsend MK, Cassidy A, Franke AA, et al. Gut Finnish men and women. J Nutr 2003;133:1830–3. microbiota metabolites of dietary lignans and risk of type 2 diabetes: a 79. Roncaglia L, Amaretti A, Raimondi S, Leonardi A, Rossi M. Role of prospective investigation in two cohorts of U.S. women. Diabetes Care bifidobacteria in the activation of the lignan secoisolariciresinol digluco- 2014;37:1287–95. side. Appl Microbiol Biotechnol 2011;92:159–68. 59. Goodman MT, Shvetsov YB, Wilkens LR, Franke AA, Le Marchand L, 80. Yoder S, Lancaster S, Hullar MAJ, Lampe JW. Gut microbial metabolism of Kakazu KK, et al. Urinary phytoestrogen excretion and postmenopausal plant lignans: influence on human health. In:Del Rio D, Tuohy K, editors. breast cancer risk: the multiethnic cohort study. Cancer Prev Res Diet–microbe interactions in the gut: Oxford, UK: Elsevier; 2014. 2009;2:887–94. 81. Lee GH, Kumar S, Lee JH, Chang DH, Kim DS, Choi SH, et al. Genome 60. Clavel T, Henderson G, Alpert CA, Philippe C, Rigottier-Gois L, Dore J, et al. sequence of Oscillibacter ruminantium strain GH1, isolated from rumen of Intestinal bacterial communities that produce active estrogen-like com- Korean native cattle. J Bacteriol 2012;194:6362. pounds enterodiol and enterolactone in humans. Appl Environ Microbiol 82. Lee GH, Rhee MS, Chang DH, Lee J, Kim S, Yoon MH, et al. Oscillibacter 2005;71:6077–85. ruminantium sp. nov., isolated from the rumen of Korean native cattle. Int 61. Clavel T, Dore J, Blaut M. Bioavailability of lignans in human subjects. Nutr J Syst Evol Microbiol 2013;63:1942–6. Res Rev 2006;19:187–96. 83. Turnbaugh PJ, Gordon JI. The core gut microbiome, energy balance and 62. Lampe JW, Atkinson C, Hullar MA. Assessing exposure to lignans and their obesity. J Physiol 2009;587:4153–8. metabolites in humans. J AOAC Int 2006;89:1174–81. 84. O'Keefe SJ. Nutrition and colonic health: the critical role of the microbiota. 63. Eeckhaut E, Struijs K, Possemiers S, Vincken JP, Keukeleire DD, Ver- Curr Opin Gastroenterol 2008;24:51–8. straete W. Metabolism of the lignan macromolecule into enterolignans 85. Wu GD, Bewtra M, Hoffmann C, Chen YY, Keilbaugh SA, Bittinger K, et al. in the gastrointestinal lumen as determined in the simulator of the Controlled feeding experiments demonstrate the impact of diet on the human intestinal microbial ecosystem. J Agric Food Chem 2008; composition of the human gut microbiome. Gastroenterology 2011;140: 56:4806–12. S47-S. 64. Wang L, Meselhy MR, LI Y, Qin G, Hattori M. Human intestinal bacteria 86. Wu GD, Chen J, Hoffmann C, Bittinger K, Chen YY, Keilbaugh SA, et al. capable of transforming secoisolariciresinol diglucoside to mammalian Linking long-term dietary patterns with gut microbial enterotypes. Science lignans, enterodiol and enterolactone. Chem Pharmol Bull 2000; 2011;334:105–8. 48:1606–10. 87. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C, et al. A human 65. Clavel T, Borrmann D, Braune A, Dore J, Blaut M. Occurrence and activity of gut microbial gene catalogue established by metagenomic sequencing. human intestinal bacteria involved in the conversion of dietary lignans. Nature 2010;464:59–65. Anaerobe 2006;12:140–7. 88. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, Ley RE, 66. Jin J-S, Kakiuchi N, Hattori M. Enantioselective oxidation of enterodiol to et al. A core gut microbiome in obese and lean twins. Nature 2009;457: enterolactone by human intestinal bacteria. Biol Pharm Bull 2007;30: 480–4. 2204–6. 89. Claesson MJ, O'Toole PW. Evaluating the latest high-throughput molecular 67. Jin JS, Hattori M. Further studies on a human intestinal Bacterium Rumi- techniques for the exploration of microbial gut communities. Gut nococcus sp END-1 for transformation of plant lignans to mammalian Microbes 2010;1:277–8. lignans. J Agric Food Chem 2009;57:7537–42. 90. Koren O, Spor A, Felin J, Fak F, Stombaugh J, Tremaroli V, et al. Human 68. De Filippo C, Cavalieri D, Di Paola M, Ramazzotti M, Poullet JB, Massart S, oral, gut, and plaque microbiota in patients with atherosclerosis. Proc Natl et al. Impact of diet in shaping gut microbiota revealed by a comparative Acad Sci U S A 2011;108 Suppl 1:4592–8. study in children from Europe and rural Africa. Proc Natl Acad Sci U S A 91. Huse SM, Ye Y, Zhou Y, Fodor AA. A core human microbiome as viewed 2010;107:14691–6. through 16S rRNA sequence clusters. PLoS ONE 2012;7:e34242. 69. Cotillard A, Kennedy SP, Kong LC, Prifti E, Pons N, Le Chatelier E, et al. 92. Human Microbiome Project C. Structure, function and diversity of the Dietary intervention impact on gut microbial gene richness. Nature healthy human microbiome. Nature 2012;486:207–14. 2013;500:585–8. 93. Rajilic-Stojanovic M, Heilig HG, Tims S, Zoetendal EG, de Vos WM. Long- 70. Chang JY, Antonopoulos DA, Kalra A, Tonelli A, Khalife WT, Schmidt term monitoring of the human intestinal microbiota composition. Envi- TM, et al. Decreased diversity of the fecal Microbiome in recurrent ron Microbiol 2012;15:1146–59. Clostridium difficile-associated diarrhea. J Infect Dis 2008;197: 94. Adlercreutz H, Honjo H, Higashi A, Fotsis T, H€am€al€ainen E, Hasegawa T, 435–8. et al. Urinary excretion of lignans and isoflavonoid phytoestrogens in 71. Adlercreutz H, Fotsis T, Bannwart C, W€ah€al€aK,M€akel€a T, Brunow G, et al. Japanese men and women consuming a traditional Japanese diet. Am J Clin Determination of urinary lignans and phytoestrogen metabolites, poten- Nutr 1991;54:1093–100. tial antiestrogens and anticarcinogens, in urine of women on various 95. Milder IEJ. Lignan intake in the Netherlands and its relation with mortality. habitual diets. J Steroid Biochem 1986;25:791–7. Helsinki, Finland: Wageningen Universiteit; 2007.

554 Cancer Epidemiol Biomarkers Prev; 24(3) March 2015 Cancer Epidemiology, Biomarkers & Prevention

Enterolignan-Producing Phenotypes Are Associated with Increased Gut Microbial Diversity and Altered Composition in Premenopausal Women in the United States

Meredith A.J. Hullar, Samuel M. Lancaster, Fei Li, et al.

Cancer Epidemiol Biomarkers Prev 2015;24:546-554. Published OnlineFirst December 26, 2014.

Updated version Access the most recent version of this article at: doi:10.1158/1055-9965.EPI-14-0262

Supplementary Access the most recent supplemental material at: Material http://cebp.aacrjournals.org/content/suppl/2014/12/27/1055-9965.EPI-14-0262.DC1

Cited articles This article cites 89 articles, 28 of which you can access for free at: http://cebp.aacrjournals.org/content/24/3/546.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://cebp.aacrjournals.org/content/24/3/546.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department Subscriptions at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cebp.aacrjournals.org/content/24/3/546. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.